Experiments on Voter Participation and Partisanship in Southern and East Africa

Session Submission Type: Full Submitted Panel

Session Description

Greater participation in elections is considered a core component of a vibrant democracy and a key mechanism for strengthening accountability of political leaders, while a strong popular mandate is often key for incumbents to claim legitimacy. In understanding citizens’ choices to participate, theorists have differentiated the ability to vote from the motivation to vote and highlighted that each is driven by radically different factors (Harder and Krosnick 2008). This panel uses four large-scale experiments in four African countries – Tanzania, Kenya, Malawi and South Africa – to cleanly identify the separate effects, holding other factors constant, of constraints on ability to register and vote and of various factors driving motivation. In some cases, we can identify how ability and motivation interact.

In considering the ability to vote, researchers considering barriers to political participation in poor African countries have focused on violence, electoral fraud, or disillusionment with democracy linked with an illegitimate electoral system. They have also examined how poor voting knowledge, often a symptom of poverty or illiteracy, can reduce political participation.

While many of these themes resonate in poor countries or newer democracies emerging from authoritarianism, they are less relevant in established African democracies or in countries nearing middle income status. Three papers explore the role of material and informational constraints in Africans’ decisions to register and vote in more established or slightly wealthier African democracies. Do transportation and spatial constraints prevent voters from participating? How do such constraints interact with limited information on how to register to vote (Harris/van der Windt)? Does understanding more complex voting systems improves participation (Orkin)? What are the effects of better understanding whether a past incumbent chose voters’ district to receive aid (Jablonski/Seim)?

In considering the motivation to vote, researchers have focused on ethnicity and clan ties or on clientilism and vote-buying. Again, we provide a fresh perspective on the motivation to vote in more established African democracies. We consider two motivations to vote: voters’ sense of political efficacy and their beliefs about whether their electoral participation has tangible effects. Voters in dominant-party democracies may have little faith in electoral outcomes driving greater accountability or in their own ability to act on the system. While they may believe that democracy is procedurally legitimate, they may doubt its substantive legitimacy, in the sense that it delivers outcomes voters care about. One intervention builds participants’ sense of political efficacy through a schools-based accountability intervention for parents and studies the effect on political participation (Lieberman/Zhou). Another varies whether voters know that an upcoming election is going to be highly contested, using the first electoral polls conducted in South Africa, and examines whether information alters participants’ beliefs that their vote can make a difference and their election-day choices (Orkin). A final paper explores the relationship between voters’ beliefs about whether incumbents have distributed aid to benefit their communities and their vote choices (Jablonski/Seim).

The experiments also use cross-cut factorial designs to examine how constraints on the ability to vote and motivations interact. Does information (Harris/van der Windt) or an improved sense of efficacy (Lieberman/Zhou) encourage voters to overcome material constraints? Does receiving information improve political efficacy (Jablonski/Seim, Orkin)? Do material constraints affect whether participants have the bandwidth or will to act on new information, such that interventions may be less effective for poorer respondents (all papers)?

The papers address these questions using large-scale quantitative field experiments and original, first-hand quantitative data collection, complemented by administrative data on voting (Orkin, Harris/Van der Windt) and qualitative focus group data (Jablonski/Seim, Lieberman/Zhou). Panellists will discuss emerging methodological debates, particularly pre-registration, as well as some innovations in measurement, such as capturing beliefs and expectations (Jablonski/Seim, Orkin), verifying turnout in poor contexts (Harris/Van der Windt, Orkin, Lieberman/Zhou), and measuring partisanship (Jablonski/Seim, Orkin). The panel discussant, de Kadt, studies economic and psychological barriers to registration, voting and political participation in Zambia and South Africa, using both experimental and administrative data, and thus brings deep methodological and substantive knowledge to the discussion.

Sub Unit

• DIVISION 53: AFRICAN POLITICS

Individual Presentations

• The Material Constraints to Participation: Experimental Evidence from Kenya - J. Andrew Harris, New York University Abu Dhabi; Peter Van der Windt, Columbia University

• Validated Participation, Efficacy, and Citizenship: Evidence from a Phase I RCT - Evan S. Lieberman, Massachusetts Institute of Technology; Yang-Yang Zhou, University of British Columbia

• Polls, Turnout and Partisanship in South Africa's First Competitive Election  - Kate Orkin, University of Oxford

• Aid Does Not Always Help Incumbents: An Experiment on Voter Behaviour in Malawi - Ryan Steele Jablonski, London School of Economics and Political Science; Brigitte Seim, University of North Carolina-Chapel Hill

Chair

• Ryan Steele Jablonski London School of Economics and Political Science

Discussant

• Daniel de Kadt UC Merced